A spinal cord injury (SCI) alters the communication between the brain and spinal cord, leading to unrecoverable neurological deficits, including impairments or even complete loss of upper-limb and lower-limb motor functions. The consequences of these chronic impairments are dramatic for the affected person, their family, and society. Currently, there is no approved therapy to improve motor recovery from SCI. Consequently, there is no approved technology available to address paralysis. We pioneered a Brain Spine Interface (BSI) that established a proportional link between motor intentions decoded from recordings of motor cortex activity and the modulation of electrical stimulation of the spinal cord. This BSI restored voluntary control of walking in a non-human primate model of leg paralysis. In this scenario, the activation of spinal circuits through the digital link coincides perfectly with the attempted activation of the same circuits by residual pathways from the brain. Therefore, there is a synergistic cooperation between natural and prosthetic control signals. This result compelled us to test this concept in humans. The ERC-POC BrainGait supported this transition.
EPFL obtained approval from the Swiss competent authorities (CER-VD2020-01814) to implant a BSI in three patients with chronic SCI (clinicaltrials.gov NCT04632290). This BSI interfaces the WIMAGINE system developed by Clinatec (CEA, France) with a repurposed neurostimulation systems (ACTIVA-RC, Medtronic). The first participant was implanted on July 28th 2021. After 2 sessions of calibration, the decoder was able to detect the intention to mobilize the hip, knee and ankle of both legs with high accuracy and high reliability. The BSI enabled the participant to perform dynamic and sustained movements of these 6 joints independently. The BSI also enabled the participant to walk independently with the most natural gait that we ever observed in people with SCI. Indeed, the participant described the control of his legs as remarkably natural, since the proportional control over the amplitude of EES enabled a perfect match between the activation of muscles and his intended movements. He also described a robust synergy between his natural attempt to activate muscles through residual pathways and the prosthetic control signals. This proof-of-concept BSI revealed that ECoG recording technologies are the best compromise between invasiveness and signal-to-noise ratio to develop a BSI for widespread dissemination. This milestone also demonstrates the relevance and feasibility of a fully-implantable BSI to restore lower-limb movements in people with paralysis due to a SCI.
